What’s the opportunity about?

LLMs are quickly becoming state-of-the-art for several NLP tasks and at Unbabel we are continuously improving our own model: Tower LLM. We are looking for a Linguist to help us test and evaluate Tower LLM in different languages and NLP tasks, such as MT and QE, Named Entities localization, and Cultural Transcreation.

Responsibilities

• Executes linguistic analysis tasks and prompt engineering following the existing Unbabel standards and practices with the guidance from more experienced Linguistic Services researchers:
  • Data collection and annotation
  • Quality processes (testing and evaluation) for Named Entity localization, Style Guides application, Cultural Transcreation
  • MT and QE benchmarkings
  • Context in MT
  • Guardrails for TowerLLM
• Participates in research activities and productization projects under direct supervision

Requirements:

• Relevant BSc degree or an in-progress MA in the areas of Linguistics, Computational Linguistics and/or Translation Studies.
• Interest in Translation Quality evaluation methodologies (metrics and quality standards)
• Basic knowledge of linguistic annotation methodologies
• High verbal and written communication in English (C1-2 CEFR Level)
• High proficiency (C1 CEFR Level) in another language other than English and the native language
• Strong communication skills
• An analytical mind with problem-solving abilities